Week 9 - Decision Theory Flashcards

1
Q

How do you track decisions?

A

Decision log
- Who
-When
- Why

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Routine Decision

A
  • small impact
  • Reversible
  • short term
  • data available
  • standard decision process
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Make-or-break decision

A

High impact
irreversible
long term
safety or product liability is critical
no well defined decision process

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Structured decision processes

A
  • Qualitative
  • Quantitative
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How is risk described?

A

A - Future Event
C - Activity consequence
C* - Prediction of C
U - Uncertainty about what value C will take
P - probability given K
K - background information

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Risk analysis methods

A
  • Simplified (Qualitative )
  • Standard (Quantitative or qualitative)
  • Model-based (primarily quantitative)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

simplified risk analysis

A

informal procedure using brainstorming and discussion

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Standard Risk analysis

A

Formalized procedure where recognized risk analysis methods are used.
Risk matrices are often used.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Model-based risk analysis

A

Makes use of techniques such as event tree and fault tree analysis.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Fault Tree Analysis (FTA)

A

System analysis techniques used to determine the root causes and probability of occurrence of a specified undesired event.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

OR gate probability

A

P(A U B) = P(A) + P(B) - P(A)P(B)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

OR gate probability for mutually exclusive events

A

P(A U B) = P(A) + P(B)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

OR gate with 3 inputs Probability

A

P(A+B+C) = P(A)+P(B)+P(C) - P(A)P(B)-P(A)P(C)-P(B)*P(C) + P(A)P(B)P(C)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

AND Gate probability

A

P(A ∩ B) = P(A)P(B)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Cut sets

A

Sets of Events that together cause the top undesired event to occur.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What does a low-order cut set indicate?

A

High safety vulnerability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Reactive FTA

A

Used after an accident as an investigation method

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Proactive FTA

A

Performed during system development to predict and prevent future problems.

19
Q

Event Tree analysis (ETA)

A

Inductive Procedure that shows all possible outcomes resulting from an accidental initiating event.

20
Q

How is ETA useful?

A
  • identifies potential accident scenarios
  • scenario frequencies
  • scenario consequences
21
Q

ETA Structure

A

initiating event
pivotal events
outcomes
probability
consequences

22
Q

For ETA:
Multiply along…
Sum across…
set of branches probability must sum to…

A

multiply along horizontal branches
sum across vertical branches
must sum to 1

23
Q

Uses of Markov Chains

A
  • Predicting traffic flows
  • communications networks load
  • genetic issues
  • currency exchange rates
  • population dynamics
24
Q

stationary matrix

A

Matrix converges to steady state
Called a Regular Markov Chain

25
Q

How to tell if a matrix is regular?

A

All entries become positive

26
Q

Absorbing state

A

If the state is impossible to leave once it is entered

27
Q

Requirements for an absorbing Markov Chain

A

Minimum of 1 absorbing state
It is possible to go from each non-absorbing state to at-least one absorbing state

28
Q

What makes a decision sensitive

A
  • After a test the decision is different for different test results.
  • The decision changes after more tests
29
Q

Reasons to not do a test for a decision

A
  • avoid costs if result is unlikely to affect decision
  • If multiple tests are available a different one may be chosen.
30
Q

Pre-posterior

A

doing a test, updating the probability distribution and deciding what to do next.

31
Q

Posterior (Bayes)

A

Involves calculating probabilities and deciding on action given an experiment has already been done. (result is known)

32
Q

Types of uncertainty

A

aleatory - random
Epistemic - Lack of Knowledge

33
Q

Type 1 Error

A

Rejecting a True Hypothesis (False Positive)

34
Q

Type 2 Error

A

Accepting a False Hypothesis (False Negative)

35
Q

Utility

A

Defined as the state of being satisfied, useful, beneficial etc…
Ranks happiness

36
Q

Marginal Utility

A

Additional utility gained from the consumption of one additional unit.

37
Q

Total utility

A

Total amount of satisfaction that a person/decision maker can receive from the consumption of all units of a good or service.

38
Q

Law of Diminishing Marginal Utility

A

As you consume more you get diminishing utility but you still consume the same number of resources.

39
Q

Risk Neutral

A

bases all decisions on expected utility

40
Q

Risk Averse

A

Will not be comfortable with a large probability of big loss

41
Q

Not Risk averse

A

Will be comfortable with a small probability of a big win

42
Q

Latin Hypercube sampling

A

Decide how many sample points to use.
Each sample point the row and column it was taken from must be recorded

43
Q

Orthogonal Sampling

A

Sample space is divided into equal subspaces.
All points are chosen simultaneously ensuring that each subspace is sampled with the same density